I have DIY assembled an Y6 multirotor running on Pixhawk. I’ve done
several test flights with AUTO mission used for photogrammetry grid
with 75/75 overlap and using A5000 camera remotely triggered by Pixhawk
using CAM trigger messages. Now normally this solution works great on
other Pixhawk based drones but this one has a particular problem: each
flight there are more CAM messages recorded in the log than the actual
number of images recorded by the camera. Now I wonder what could be the
problem source… The only thing that came to my mind is that Pixhawk
actually triggers CAM messages (using CAM_TRIGG_DIST command) even
before the first waypoint of AUTO mission is reached. This could happen
when the AUTO mode has been triggered before the first waypoint is
reached. But I have no way of how to find out if this is true or false
because I cannot geotag images (CAM vs. Image numbers don’t match). I
always get lower number of images from the camera than the number of CAM
messages. Can you please confirm my suspicion ? Logs here: https://drive.google.com/open?id=0B0VZlBaoFwO-czQ2MHhDcEZzTDA https://drive.google.com/open?id=0B0VZlBaoFwO-cW1IcFJHRjNubDg https://drive.google.com/open?id=0B0VZlBaoFwO-cmE4ZHNEWWJiTjA
The time between the images is 1s so this should not be a problem for the camera.
For a start 1s between images on a a5000 is pretty close to the limit.
We have used a5000 and a6000 for this job and limit it to 1.5s between shots for safety.
At 5m/s that usually equates to 9m between images.
The SD card quality is critical or you will get a bottleneck effect happening and lose shots while it processes a bunch of photos.
The auto focus is another real slow down and can miss shots for you.
In missions it will start to trigger whenever it is auto and sees a movement greater than trigger distance, but it still should match between log and shots.
If you open your log as a spreadsheet and filter by CAM then compare to the directory listing of shots you will soon see where you are missing them.
Thanks for the suggestion. I will test slower mission speed if this helps. I tested AF and MF modes but both have missing shots. So maybe it really is the problem of the camera speed not keeping up…
@Thorsten For safety we keep time between shots at 1.5s which has been jpg in the past but at present we are trying RAW.
The SD car MUST be the fastest possible or it will bottle neck.
We typically do ~500 shot runs and occasionally get one or two dropped photos.
The copter I am setting up now will be using hotshoe feedback for the cam message on a Septentrio board if we can stop the Pixhawk FW from crashing at random intervals.